Journal of Membrane Science, Vol.473, 318-326, 2015
Characterization of hollow fiber membranes by impedance spectroscopy
Electrical impedance spectroscopy (EIS) is a simple and non-destructive method to characterize membranes and to monitor membrane processes. It can be used for instance to determine layer thicknesses or to observe fouling. Up to now it has only been applied to flat sheet membranes using a four-electrode measuring technique. Hollow fiber membranes are finding increasing application but monitoring the state of such membranes during operation is a difficult task. The aim of our work presented here is to adapt the electrical impedance spectroscopy method for use with hollow fiber membranes. For this purpose a new membrane module has been developed which allows electrical impedance measurements to be made on hollow fibers and capillary membranes using a 2-terminal method. For this one wire-shaped electrode is located inside the hollow fiber membrane and one ring-shaped electrode is located outside, around the membrane. In the experiments the applied AC field used in the impedance measurements was in a radial direction; across the membrane from the lumen side to the shell side while the system was bathed by an electrolyte solution. A porous hydrophobic PP membrane with a pore size of 0.2 mu m has been intensively analyzed with the new device to explore the characteristics of the new methodology. The impedance of the ionic double layers and the frequency dependence of those impedances, at the two electrodes used in these experiments pose a major problem. These complications can normally be avoided by the use of 4-terminal methods but this is even more intricate to do with the hollow fiber geometry due to lack of space. Nevertheless, using the 2-terminal technique, it was possible to obtain meaningful impedance spectra. The membrane's impedance could be determined and a significant influence of the membrane wetting could be observed. Such wetting phenomena are often speculated about when analyzing flux data, but can now be quantified systematically. Furthermore, we present a comprehensive model to illustrate the opportunities and challenges for the developed technique. (C) 2014 Elsevier B.V. All rights reserved.